MMLU-by-task - Miscellaneous — leaderboard

Metric: Accuracy (%). Source: huggingface.co. 1257 models tracked.

Top models

#ModelScore
1falcon-180B87.61
2Llama 2 70B Base86.21
3StableBeluga286.08
4Llama 2 70B Chat85.95
5Llama 2 70B Chat (HF)82.76
6Mistral-7B-v0.181.74
7LLaMA-65B81.61
8Llama 2 70B Chat GPTQ80.97
9LLaMA-30B78.16
10vicuna-33B-v1.377.39
11MythoMax-L2-13B76.88
12internlm-20B Chat75.61
13vicuna-13B-v1.574.97
14OpenHermes-13B74.97
15Llama 2 13B Chat Base74.71

Interactive version: theaggregate.ai/benchmark?slug=mmlu-by-task-miscellaneous · How the rankings work · Data refreshed daily, snapshot 2026-07-22.